ornith

Run Ornith, Qwen, Nemotron & DeepSeek optimized on AMD Strix Halo — unified OpenAI-compatible server with tuned ROCmFP4 quants, validated 262K context, vision support & hot-swap model zoo (iGPU + NPU)

A standalone local-only Python MCP server that gives any IDE persistent, workspace-isolated memory. works with any IDE supporting MCP servers

Ornith 1.5 Dropped: This 9B AI Coder BEATS Qwen 3.8? - Offline AI code generation scaffold using Ollama GGUF models.

A Python CLI coding agent powered by remote Ollama, with workspace inspection and command-execution tools over Tailscale.

A fully local OpenCode + llama.cpp + Ornith 1.5 coding-agent setup tuned for 128K context on a tight 6 GB VRAM budget.

Privacy-first guide and helper scripts for running Pi with local llama.cpp models on Apple Silicon.

Research & experiments running open-source LLMs locally on an 8GB VRAM laptop using Ollama/LM Studio with LangChain integration and performance benchmarks

Open models extended to 1M context with YaRN and certified needle by needle: Ornith, Gemma 4 uncensored, Qwen3.6 uncensored. MTP speculative decoding grafts, vision, full test harness.